基于干扰观测器的轧机扭振抑制策略研究

Torsional Vibration Suppression Strategy for Rolling Mill Based on Disturbance Observer

考虑异步电机转子偏心,建立机电耦合模型,将转子偏心视为系统内扰,状态观测器与干扰观测器相结合,利用稳定裕度调整观测器参数,实现系统内环的干扰观测及补偿。同时考虑到系统复杂、参数不易测量的特点,采用神经网络速度控制器,在线更新权重系数,实现速度在线控制。通过仿真实验,验证了所提出的控制结构在轧机扭振抑制方面的有效性。

In view of eccentricity of asynchronous motor, an eleetromeehanieal coupling model was established, with the rotor eccentricity regarded as a system disturbance and the state observer combined with the disturbance observer. And the interferenee observations and compensation of the inner ring of the system can be realized by using stability margin to adjust the observer parameters. Meanwhile, in consideration of the complexity of system and difficulty in the parameter measurement, a neural network speed controller was adopted to online update weight coefficient, so as to achieve speed control online. A simulation experiment has verified the effectiveness of the proposed control structure in torsional vibration suppression of the rolling mill.